Objective 1 was to develop a method for PBM that is computationally efficient and compute the results with higher precision. This allows the researcher to have an advanced prediction of how the system behaves when the process parameters are varied, hence further leads to less powder wastage which was the main target of the project. The mathematical model developed for continuous manufacturing unit (twin-screw granulator) using population balance model (PBM) approach by incorporating the process parameters in PBM. For solving these complex models, an accurate and efficient numerical technique is required. Therefore, we developed two finite volume methods to solve a simultaneous aggregation-breakage PBM.
The second objective was to extract the mean residence time (MRT) so that during the granulation we have more control over the process and desired quality granules can be manufactured. This highly depends on the knowledge of MRT which describes the stay of powder in a specific part of the screw. Different properties granules can be prepared by correlating the MRT with process parameters such as liquid to solid ratio (L/S), powder feed rate, and screw speed. This gives a better understanding of the behavior of granules prepared during the granulation process. In addition, the most important parameter while using the wet TSG is the MRT of powder inside the barrel. Process parameters including feed flow rate, screw speed, and L/S are correlated with the obtained values of MRT to build a predictive tool. Artificial neural network (ANN) modeling is implemented to predict the MRT of pharmaceutical formulation in a wet TSG.
Next, we develop a model that has the ability to predict mechanistically the behaviour of particles inside the TSG. Experimental data is collected for Microcrystalline cellulose (MCC-101, Avicel pH 101) was granulated with water in a TSG. A five compartmental population balance model (CPBM) is developed and 10 parameters related to aggregation and breakage PBM is optimized. In addition, kriging interpolation is used to interpolate for new values of empirical parameters at different L/S and screw speeds. This model has the tendency to extract new data and further assists in reducing the waste of the powder in the pharmaceutical industry. Five CPBM is developed for the TSG. Moreover, Kriging interpolation is used to interpolate for new values of empirical parameters at different L/S and screw speeds. Finally, the CPBM model is calibrated and validated using the experimental data.
